Add MultiFrameDataset

Retrieves video sequence patches rather than single images.

codes/data/multi_frame_dataset.py

@@ -0,0 +1,89 @@
+from data.base_unsupervised_image_dataset import BaseUnsupervisedImageDataset
+import numpy as np
+import torch
+from bisect import bisect_left
+import os.path as osp
+
+class MultiFrameDataset(BaseUnsupervisedImageDataset):
+    def __init__(self, opt):
+        super(MultiFrameDataset, self).__init__(opt)
+        self.num_frames = opt['num_frames']
+
+    def chunk_name(self, i):
+        return osp.basename(self.chunks[i].path)
+
+    def get_sequential_image_paths_from(self, chunk_index, chunk_offset):
+        im_name = self.chunk_name(chunk_index)
+        source_name = im_name[:-12]
+        frames_needed = self.num_frames - 1
+        # Search backwards for the frames needed. We are assuming that every video in the dataset has at least frames_needed frames.
+        search_idx = chunk_index-1
+        while frames_needed > 0:
+            if source_name in self.chunk_name(search_idx):
+                frames_needed -= 1
+                search_idx -= 1
+            else:
+                break
+
+        # Now build num_frames starting from search_idx.
+        hqs, refs, masks, centers = [], [], [], []
+        for i in range(self.num_frames):
+            h, r, c, m, p = self.chunks[search_idx + i][chunk_offset]
+            hqs.append(h)
+            refs.append(r)
+            masks.append(m)
+            centers.append(c)
+            path = p
+        return hqs, refs, masks, centers, path
+
+    def __getitem__(self, item):
+        chunk_ind = bisect_left(self.starting_indices, item)
+        chunk_ind = chunk_ind if chunk_ind < len(self.starting_indices) and self.starting_indices[chunk_ind] == item else chunk_ind-1
+        hqs, refs, masks, centers, path = self.get_sequential_image_paths_from(chunk_ind, item-self.starting_indices[chunk_ind])
+
+        hs, hrs, hms, hcs = self.resize_hq(hqs, refs, masks, centers)
+        ls, lrs, lms, lcs = self.synthesize_lq(hs, hrs, hms, hcs)
+
+        # Convert to torch tensor
+        hq = torch.from_numpy(np.ascontiguousarray(np.transpose(np.stack(hs), (0, 3, 1, 2)))).float()
+        hq_ref = torch.from_numpy(np.ascontiguousarray(np.transpose(np.stack(hrs), (0, 3, 1, 2)))).float()
+        hq_mask = torch.from_numpy(np.ascontiguousarray(np.stack(hms))).unsqueeze(dim=1)
+        hq_ref = torch.cat([hq_ref, hq_mask], dim=1)
+        lq = torch.from_numpy(np.ascontiguousarray(np.transpose(np.stack(ls), (0, 3, 1, 2)))).float()
+        lq_ref = torch.from_numpy(np.ascontiguousarray(np.transpose(np.stack(lrs), (0, 3, 1, 2)))).float()
+        lq_mask = torch.from_numpy(np.ascontiguousarray(np.stack(lms))).unsqueeze(dim=1)
+        lq_ref = torch.cat([lq_ref, lq_mask], dim=1)
+
+        return {'LQ': lq, 'GT': hq, 'gt_fullsize_ref': hq_ref, 'lq_fullsize_ref': lq_ref,
+             'lq_center': torch.tensor(lcs, dtype=torch.long), 'gt_center': torch.tensor(hcs, dtype=torch.long),
+             'LQ_path': path, 'GT_path': path}
+
+
+if __name__ == '__main__':
+    opt = {
+        'name': 'amalgam',
+        'paths': ['F:\\4k6k\\datasets\\ns_images\\vixen\\full_video_256_tiled_with_ref'],
+        'weights': [1],
+        'target_size': 128,
+        'force_multiple': 32,
+        'scale': 2,
+        'eval': False,
+        'fixed_corruptions': [],
+        'random_corruptions': ['color_quantization', 'gaussian_blur', 'motion_blur', 'smooth_blur', 'noise', 'saturation'],
+        'num_corrupts_per_image': 1,
+        'num_frames': 10
+    }
+
+    ds = MultiFrameDataset(opt)
+    import os
+    os.makedirs("debug", exist_ok=True)
+    for i in range(100000, len(ds)):
+        import random
+        o = ds[random.randint(0, 1000000)]
+        k = 'gt_fullsize_ref'
+        v = o[k]
+        if 'path' not in k and 'center' not in k:
+            fr, f, h, w = v.shape
+            for j in range(fr):
+                import torchvision
+                torchvision.utils.save_image(v[j].unsqueeze(0), "debug/%i_%s_%i.png" % (i, k, j))